The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
In Content Based Image Retrieval (CBIR) system, Relevance Feedback (RF) technique can substantially promote retrieval accuracy. As a popular algorithm of Relevance Feedback, Query movement was widely applied in this area and it moves the query in the input space towards relevant images. However, traditional Query Movement algorithm has significant disadvantages. It is to reduce the similarity between...
Local boundary information plays an important role in shape description. In this paper, an improved Arc-Height Function is presented for shape description and retrieval. The Improved Arc-Height is independent of centroid distance (CD) and can capture local boundary information very accurately. And an effective fusion strategy is utilized to integrate Fourier features derived from the improved Arc-Height...
Salient region-based image retrieval is one of the hotspots in the domain of content-based image retrieval; however the metrics about region saliency is not in the uniform frame. The research on visual attention has shown that the factors including color, texture, scale and position influence on visual perception mostly. Consequently, the algorithm of salient region extraction is proposed by using...
In this paper, we propose an effective frame of shot segmentation in compress-domain videos. Firstly, we extract DC coefficients of I-frames in compressed-domain, and construct a sequence of DC-images which are on the basis of DC coefficients in I-Frames. Secondly, the difference between two adjacent DC-images is calculated by integrating grid-mapping dynamic window, color moments and spatial distribution...
Binarization of plate image is one of the major tasks in license plate recognition system (LPR). Many researchers have established various kinds of theories to reduce the degradation effects on a car license plate. However, most of these theories are not efficient in practice due to their expensive computational cost. This paper focuses on the binarization task of poor illumination in a practical...
In this paper, we propose the directional neighborhood distance and all-directional neighborhood distance (ADND) to measure the gray value variation between pixels in specific and all direction. The all-directional neighborhood distance is used to fuse multi-source images, and our experiments show that the proposed fusion method is effective in term of some objective evaluation indexes, such as spatial...
Double compression is quite common due to image forgery and specific steganographic algorithms. It would notably impact results of steganalysis if not treated properly. This paper discusses the effects of double compression to steganalysis. Then, we evaluate the effect of double compression using L-GEM based RBFNN in comparison to widely adopted SVMs.
This paper presents a novel and efficient decision tree construction approach based on C4.5. C4.S constructs decision tree with information gain ratio and deals with missing values or noise. ID3 and its improvement, C4.5, both select one attribute as the splitting criterion each time during constructing decision tree, adopting one step forward. Comparing with one step forward, the proposed algorithm,...
A sparse algorithm, based on empirical feature selection, is investigated from the viewpoint of learning theory. It is a novel way to realize sparse empirical feature-based learning different from the regularized kernel projection machines. Représenter theorem and error analysis of this algorithm are established without sparsity assumption of regression function. An empirical study verifies our theoretical...
Support vector machines (SVMs) are probably the most well-known models based on kernel substitution. Based on orthogonal Legendre polynomials, an orthogonal Legendre kernel function for support vector machine is proposed using the properties of kernel functions. We then prove that it satisfies the Mercer condition. Compared to traditional kernel functions such as polynomial or gaussian kernels, orthogonal...
This paper presents an algorithm based on a combination of Discrete Wavelet Transforms and back-propagation neural networks for location of interturn faults in a two-winding three-phase transformer. Fault conditions of the transformer are simulated using ATP/EMTP in order to obtain current signals. The training process for the neural network and fault diagnosis decision are implemented by MATLAB....
This paper focuses on fault diagnosis using neuro-fuzzy network. It is shown that less reliable result may be derived as the network takes no consideration of previous state information in online fault diagnosis. To solve this problem, we combine a modified neuro-fuzzy network with the evidence update theory. Besides, a new updating rule that combining the Jeffery-like rule and linear combination...
Steganalysis is detecting and decoding hidden data within a given media and is taken as a countermeasure to steganography. There has been quite some effort in audio steganalysis for additive embedding model. But, results are disappointing when they distinguish the cover-audio signal with multiplicative noise and the stego-audio signal. In this paper, multiplicative noise is changed to additive noise...
Nowadays, the mobile phones are becoming part of our daily life. They transformed from communication devices to personal computers. They are capable of storing personal information and connect to other applications where security is required. The need for securing the confidential information and the other application is increasing every day. Fingerprint authentication has been used in many different...
This paper proposes an approach integrating 3D Bayesian level set method with volume rendering for brain tumor and tissue segmentation and rendering. A prior probability estimation of the tumor and tissue is incorporated into 3D Bayesian level set method for 3D segmentation. The 3D Bayesian level set method is then used to continuously segment the 3D targets (e.g., tumor, tissue, and whole brain)...
Most existing cancelable biometrie frameworks are based on one-dimensional (ID) vectors rather than two-dimensional (2D) images or feature matrices. 2D cancelable biometrics, generated directly from images of feature matrices, were proposed based on two-directional two-dimensional fusion sparse random projection ((2D)2FSRP) and two-directional two-dimensional plus sparse random projection ((2D)2PSRP),...
Individual recognition is the technique which recognizes person's identity through his gait. Gait energy image (GEI) is a classical gait representation and it can be decomposed into structural part and detailed part. Then virtual gait energy image (VGEI) can be constructed in virtual space by integrating those two different parts. The generalized principal component analysis (GPCA) is applied to VGEI...
A novel partially occluded face recognition method based on biomimetic pattern recognition (BPR), proposed by academician Shoujue Wang, is proposed in this paper. A model for partially occluded face recognition based on BPR was introduced. An experiment based on biomimetic pattern recognition adopting a PCA and LDA feature extraction method is performed. Experimental results on Yale image database...
As is well known, good segmentation is one reason for high accuracy of character recognition; this paper proposes and investigates a new technique for segmentation of handwritten Arabic scripts. A new Arabic heuristic segmenter (AHS) has been implemented. The AHS employs three new features to locate a Prospected Segmentation Point (PSP) based on shape of the word image, first, remove the punctuation...
In classification problem, single classifier may not fully catch the dataset's information. Thus, an ensemble method based on Support Vector Machine (SVM) is proposed in this paper for image scene classification. First, Scale Invariant Feature Transform (SIFT) is used to extract the features of the images, and the SIFT features are clustered to form a visual vocabulary. Then, the SIFT features of...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.